Deep Web sources contain a large of high-quality and query-related structured date. One of the challenges in the Deep Web is extracting result schemas of Deep Web sources. To address this challenge, this paper describ...Deep Web sources contain a large of high-quality and query-related structured date. One of the challenges in the Deep Web is extracting result schemas of Deep Web sources. To address this challenge, this paper describes a novel approach that extracts both result data and the result schema of a Web database. The approach first models the query interface of a Deep Web source and fills in it with a specifically query instance. Then the result pages of the Deep Web sources are formatted in the tree structure to retrieve subtrees that contain elements of the query instance, Next, result schema of the Deep Web source is extracted by matching the subtree' nodes with the query instance, in which, a two-phase schema extraction method is adopted for obtaining more accurate result schema. Finally, experiments on real Deep Web sources show the utility of our approach, which provides a high precision and recall.展开更多
Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst cas...Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst case of algorithms by such a technique. In this paper, in terms of non-functional testing, we re-define the worst case of some algorithms, respectively. By using genetic algorithms (GAs), we illustrate the strategies corresponding to each type of instances. We here adopt three problems for examples;the sorting problem, the 0/1 knapsack problem (0/1KP), and the travelling salesperson problem (TSP). In some algorithms solving these problems, we could find the worst-case instances successfully;the successfulness of the result is based on a statistical approach and comparison to the results by using the random testing. Our tried examples introduce informative guidelines to the use of genetic algorithms in generating the worst-case instance, which is defined in the aspect of algorithm performance.展开更多
This paper proposes a checking method based on mutual instances and discusses three key problems in the method: how to deal with mistakes in the mutual instances and how to deal with too many or too few mutual instan...This paper proposes a checking method based on mutual instances and discusses three key problems in the method: how to deal with mistakes in the mutual instances and how to deal with too many or too few mutual instances. It provides the checking based on the weighted mutual instances considering fault tolerance, gives a way to partition the large-scale mutual instances, and proposes a process greatly reducing the manual annotation work to get more mutual instances. Intension annotation that improves the checking method is also discussed. The method is practical and effective to check subsumption relations between concept queries in different ontologies based on mutual instances.展开更多
HTTP Adaptive Streaming(HAS)of video content is becoming an undivided part of the Internet and accounts for most of today’s network traffic.Video compression technology plays a vital role in efficiently utilizing net...HTTP Adaptive Streaming(HAS)of video content is becoming an undivided part of the Internet and accounts for most of today’s network traffic.Video compression technology plays a vital role in efficiently utilizing network channels,but encoding videos into multiple representations with selected encoding parameters is a significant challenge.However,video encoding is a computationally intensive and time-consuming operation that requires high-performance resources provided by on-premise infrastructures or public clouds.In turn,the public clouds,such as Amazon elastic compute cloud(EC2),provide hundreds of computing instances optimized for different purposes and clients’budgets.Thus,there is a need for algorithms and methods for optimized computing instance selection for specific tasks such as video encoding and transcoding operations.Additionally,the encoding speed directly depends on the selected encoding parameters and the complexity characteristics of video content.In this paper,we first benchmarked the video encoding performance of Amazon EC2 spot instances using multiple×264 codec encoding parameters and video sequences of varying complexity.Then,we proposed a novel fast approach to optimize Amazon EC2 spot instances and minimize video encoding costs.Furthermore,we evaluated how the optimized selection of EC2 spot instances can affect the encoding cost.The results show that our approach,on average,can reduce the encoding costs by at least 15.8%and up to 47.8%when compared to a random selection of EC2 spot instances.展开更多
Network functions virtualization(NFV) increases network flexibility and scalability by virtualizing network functions running on the general servers and opens the network innovations by outsourcing VNF instances in 5G...Network functions virtualization(NFV) increases network flexibility and scalability by virtualizing network functions running on the general servers and opens the network innovations by outsourcing VNF instances in 5G networks.However,it leads to the incompatibility issue among different VNF instances,which makes operators difficult to determine which VNF instances to select for Service Function Chains(SFCs).In this paper,we divide VNF instances with high compatibility into clusters used for combining VNF instances in 5G networks.Firstly,we define compatibility among different VNF instances.Secondly,aiming to maximize compatibility of each cluster,we propose a novel hypergraph clustering model that divides the VNF instances into multiple clusters.Then,the hypergraph clustering model is transformed to an evolutionary game.Thus,the cluster establishing is transformed to the game equilibrium searching.Furthermore,we propose a discrete time high order replicator dynamic algorithm to find the game equilibrium.Finally,the simulation results show that the proposed approach can improve the quality of SFCs.展开更多
We propose and discuss a novel concept of robust set stabilization by permissible controls; this concept is helpful when dealing with both a priori information of model parameters and different permissible controls in...We propose and discuss a novel concept of robust set stabilization by permissible controls; this concept is helpful when dealing with both a priori information of model parameters and different permissible controls including quantum measurements. Both controllability and stabilization can be regarded as the special case of the novel concept. An instance is presented for a kind of uncertain open quantum systems to further justify this gen- eralized concept. It is underlined that a new type of hybrid control based on periodically perturbed projective measurements can be the permissible control of uncertain open quantum systems when perturbed projective measurements are available. The sufficient conditions are given for the robust set stabilization of uncertain quantum open systems by the hybrid control, and the design of the hybrid control is reduced to selecting the period of measurements.展开更多
Many international brands have a phenomenal Chinese name which,paradoxically,comes from a rather prosaic name.The reason for this may lie in the fact that they need an outstanding translation of their names in order t...Many international brands have a phenomenal Chinese name which,paradoxically,comes from a rather prosaic name.The reason for this may lie in the fact that they need an outstanding translation of their names in order to be successful in international marketing.Hence the translation of brand names is an important part of the advertisement.And a good translation is expected to bridge the differences of cultures,languages,spending habits,thinking patterns,etc.展开更多
针对单目3D目标检测在视角变化引起的物体大小变化以及物体遮挡等情况下效果不佳的问题,提出一种融合深度信息和实例分割掩码的新型单目3D目标检测方法。首先,通过深度-掩码注意力融合(DMAF)模块,将深度信息与实例分割掩码结合,以提供...针对单目3D目标检测在视角变化引起的物体大小变化以及物体遮挡等情况下效果不佳的问题,提出一种融合深度信息和实例分割掩码的新型单目3D目标检测方法。首先,通过深度-掩码注意力融合(DMAF)模块,将深度信息与实例分割掩码结合,以提供更准确的物体边界;其次,引入动态卷积,并利用DMAF模块得到的融合特征引导动态卷积核的生成,以处理不同尺度的物体;再次,在损失函数中引入2D-3D边界框一致性损失函数,调整预测的3D边界框与对应的2D检测框高度一致,以提高实例分割和3D目标检测任务的效果;最后,通过消融实验验证该方法的有效性,并在KITTI测试集上对该方法进行验证。实验结果表明,与仅使用深度估计图和实例分割掩码的方法相比,在中等难度下对车辆类别检测的平均精度提高了6.36个百分点,且3D目标检测和鸟瞰图目标检测任务的效果均优于D4LCN(Depth-guided Dynamic-Depthwise-Dilated Local Convolutional Network)、M3D-RPN(Monocular 3D Region Proposal Network)等对比方法。展开更多
基金Supported by the National Natural Science Foundation of China (60673139, 60473073, 60573090)
文摘Deep Web sources contain a large of high-quality and query-related structured date. One of the challenges in the Deep Web is extracting result schemas of Deep Web sources. To address this challenge, this paper describes a novel approach that extracts both result data and the result schema of a Web database. The approach first models the query interface of a Deep Web source and fills in it with a specifically query instance. Then the result pages of the Deep Web sources are formatted in the tree structure to retrieve subtrees that contain elements of the query instance, Next, result schema of the Deep Web source is extracted by matching the subtree' nodes with the query instance, in which, a two-phase schema extraction method is adopted for obtaining more accurate result schema. Finally, experiments on real Deep Web sources show the utility of our approach, which provides a high precision and recall.
文摘Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst case of algorithms by such a technique. In this paper, in terms of non-functional testing, we re-define the worst case of some algorithms, respectively. By using genetic algorithms (GAs), we illustrate the strategies corresponding to each type of instances. We here adopt three problems for examples;the sorting problem, the 0/1 knapsack problem (0/1KP), and the travelling salesperson problem (TSP). In some algorithms solving these problems, we could find the worst-case instances successfully;the successfulness of the result is based on a statistical approach and comparison to the results by using the random testing. Our tried examples introduce informative guidelines to the use of genetic algorithms in generating the worst-case instance, which is defined in the aspect of algorithm performance.
基金Supported by the National Natural Sciences Foundation of China(60373066 ,60425206 ,90412003) , National Grand Fundamental Research 973 Pro-gramof China(2002CB312000) , National Research Foundation for the Doctoral Pro-gramof Higher Education of China (20020286004)
文摘This paper proposes a checking method based on mutual instances and discusses three key problems in the method: how to deal with mistakes in the mutual instances and how to deal with too many or too few mutual instances. It provides the checking based on the weighted mutual instances considering fault tolerance, gives a way to partition the large-scale mutual instances, and proposes a process greatly reducing the manual annotation work to get more mutual instances. Intension annotation that improves the checking method is also discussed. The method is practical and effective to check subsumption relations between concept queries in different ontologies based on mutual instances.
基金This work has been supported in part by the Austrian Research Promotion Agency(FFG)under the APOLLO and Karnten Fog project.
文摘HTTP Adaptive Streaming(HAS)of video content is becoming an undivided part of the Internet and accounts for most of today’s network traffic.Video compression technology plays a vital role in efficiently utilizing network channels,but encoding videos into multiple representations with selected encoding parameters is a significant challenge.However,video encoding is a computationally intensive and time-consuming operation that requires high-performance resources provided by on-premise infrastructures or public clouds.In turn,the public clouds,such as Amazon elastic compute cloud(EC2),provide hundreds of computing instances optimized for different purposes and clients’budgets.Thus,there is a need for algorithms and methods for optimized computing instance selection for specific tasks such as video encoding and transcoding operations.Additionally,the encoding speed directly depends on the selected encoding parameters and the complexity characteristics of video content.In this paper,we first benchmarked the video encoding performance of Amazon EC2 spot instances using multiple×264 codec encoding parameters and video sequences of varying complexity.Then,we proposed a novel fast approach to optimize Amazon EC2 spot instances and minimize video encoding costs.Furthermore,we evaluated how the optimized selection of EC2 spot instances can affect the encoding cost.The results show that our approach,on average,can reduce the encoding costs by at least 15.8%and up to 47.8%when compared to a random selection of EC2 spot instances.
基金supported by The National High Technology Research and Development Program of China(863)(Grant No.2014AA01A701,2015AA01A706)
文摘Network functions virtualization(NFV) increases network flexibility and scalability by virtualizing network functions running on the general servers and opens the network innovations by outsourcing VNF instances in 5G networks.However,it leads to the incompatibility issue among different VNF instances,which makes operators difficult to determine which VNF instances to select for Service Function Chains(SFCs).In this paper,we divide VNF instances with high compatibility into clusters used for combining VNF instances in 5G networks.Firstly,we define compatibility among different VNF instances.Secondly,aiming to maximize compatibility of each cluster,we propose a novel hypergraph clustering model that divides the VNF instances into multiple clusters.Then,the hypergraph clustering model is transformed to an evolutionary game.Thus,the cluster establishing is transformed to the game equilibrium searching.Furthermore,we propose a discrete time high order replicator dynamic algorithm to find the game equilibrium.Finally,the simulation results show that the proposed approach can improve the quality of SFCs.
基金Supported by the National Natural Science Foundation of China under Grant Nos 61673389,61273202 and 61134008
文摘We propose and discuss a novel concept of robust set stabilization by permissible controls; this concept is helpful when dealing with both a priori information of model parameters and different permissible controls including quantum measurements. Both controllability and stabilization can be regarded as the special case of the novel concept. An instance is presented for a kind of uncertain open quantum systems to further justify this gen- eralized concept. It is underlined that a new type of hybrid control based on periodically perturbed projective measurements can be the permissible control of uncertain open quantum systems when perturbed projective measurements are available. The sufficient conditions are given for the robust set stabilization of uncertain quantum open systems by the hybrid control, and the design of the hybrid control is reduced to selecting the period of measurements.
文摘Many international brands have a phenomenal Chinese name which,paradoxically,comes from a rather prosaic name.The reason for this may lie in the fact that they need an outstanding translation of their names in order to be successful in international marketing.Hence the translation of brand names is an important part of the advertisement.And a good translation is expected to bridge the differences of cultures,languages,spending habits,thinking patterns,etc.
文摘针对单目3D目标检测在视角变化引起的物体大小变化以及物体遮挡等情况下效果不佳的问题,提出一种融合深度信息和实例分割掩码的新型单目3D目标检测方法。首先,通过深度-掩码注意力融合(DMAF)模块,将深度信息与实例分割掩码结合,以提供更准确的物体边界;其次,引入动态卷积,并利用DMAF模块得到的融合特征引导动态卷积核的生成,以处理不同尺度的物体;再次,在损失函数中引入2D-3D边界框一致性损失函数,调整预测的3D边界框与对应的2D检测框高度一致,以提高实例分割和3D目标检测任务的效果;最后,通过消融实验验证该方法的有效性,并在KITTI测试集上对该方法进行验证。实验结果表明,与仅使用深度估计图和实例分割掩码的方法相比,在中等难度下对车辆类别检测的平均精度提高了6.36个百分点,且3D目标检测和鸟瞰图目标检测任务的效果均优于D4LCN(Depth-guided Dynamic-Depthwise-Dilated Local Convolutional Network)、M3D-RPN(Monocular 3D Region Proposal Network)等对比方法。